Re: [PATCH] Use optimized single-datum tuplesort in ExecSort

Em qua., 14 de jul. de 2021 às 22:22, David Rowley <dgrowleyml(at)gmail(dot)com>
escreveu:

> On Thu, 15 Jul 2021 at 12:30, Ranier Vilela <ranier(dot)vf(at)gmail(dot)com> wrote:
> >
> > Em qua., 14 de jul. de 2021 às 21:21, David Rowley <dgrowleyml(at)gmail(dot)com>
> escreveu:
> >> But, in v8 there is no additional branch, so no branch to mispredict.
> >> I don't really see how your explanation fits.
> >
> > In v8 the branch occurs at :
> > + if (ExecGetResultType(outerPlanState(sortstate))->natts == 1)
>
> You do know that branch is in a function that's only executed once
> during executor initialization, right?
>
There's a real difference between v8 and v6, if I understood correctly.

v6 the branches is per tuple:
+ if (tupDesc->natts == 1)

v8 the branches is per state:
+ if (ExecGetResultType(outerPlanState(sortstate))->natts == 1)

I think that a big different way to solve the problem.
Or am I getting it wrong?

If the sortstate number of attributes is equal to 1, is it worth the same
for each tuple?
Can you explain this, please?

regards,
Ranier Vilela